AI Briefing
KO

GPT-5.5 System Card

·2026.04.23 20:00

Key point

It summarizes GPT-5.5's scope of use, safety evaluations, and new guardrails.

Details

GPT-5.5 is a new model tailored for code writing, online research, information analysis, creating documents and spreadsheets, and tasks that span multiple tools. Compared to previous models, it grasps tasks faster, uses tools more effectively with less guidance, and checks its own results to see tasks through to completion.

Before release, it underwent a full safety evaluation including the Preparedness Framework, along with targeted red-teaming for advanced cybersecurity and biology capabilities. It also incorporated real-world usage feedback from nearly 200 early access partners prior to launch, applying the strongest level of guardrails to reduce misuse while preserving legitimate advanced use cases.

GPT-5.5 Pro operates as a configuration that uses parallel test-time compute on the same base model, and was evaluated separately in cases where the associated risks or safeguards could differ. Unless otherwise noted, the results presented in the document are from evaluations conducted in an offline environment.

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.